Of the 510 stroke patients who had been admitted to the rehabilitation unit of a university hospital in Seoul within the period of August 2005 to June 2008, and who had received comprehensive rehabilitation treatment, 188 were included in this study, and those who had a history of surgery of the urinary system, who had been diagnosed with diabetes and Parkinson's disease, who had been taking autonomic-nervous-system drugs, or who had been taking antibiotics at the time of admission or transfer to the rehabilitation unit were excluded. All the patients in this study can self void at the time of admission to the rehabilitation unit (male: n=105; female: n=83; mean age: 67.1 years).

This was a medical-recordbased based retrospective study on stroke patients who had received comprehensive inpatient rehabilitation treatment. For all the patients in this study who had been admitted to the rehabilitation unit, the PVR was measured, using a bladder scanner Biocon-500™ (Mcube Technology, Seoul, Korea), once daily for the first 72 hours (total of three times) after admission or transfer to the rehabilitation unit. The measurement was performed by skillful nurses from the rehabilitation unit. The measurement data were recorded on the medical chart, and the mean measurements were compared. For the urine test, the midstream urine was collected, and if a patient could not express his/her voiding sense due to decreased cognitive function, urine was collected using the aseptic catheterization. A routine urine test was performed once a week from the time of admission or transfer to the rehabilitation unit to the time of discharge therefrom, and additional tests were performed if symptoms suggestive of UTI developed during the week. In addition, a urine culture test was performed at the time of admission or transfer to the rehabilitation unit and when symptoms and signs of UTI were observed. In the present study, UTI was defined as (i) the presence of fever and of symptoms of UTI (urodynia, urinary frequency, and urinary urgency) and (ii) the presence of bacteriuria (>10⁵/ml or >10³/ml bacterial count) and pyuria (>10/mm³ white blood cell count).9,10 The urine volume was measured; and a urinary test as well as the Korean Mini-Mental State Examination (K-MMSE) and Korean-Modified Barthel Index (K-MBI) were performed. The number of days from the onset of the stroke to the admission to the rehabilitation unit, and the number of days of Foley catheter indwelling since the onset of stroke were examined. The stroke types and locations were classified according to the data obtained from the brain computed tomography and magnetic resonance imaging performed during the acute phase of the stroke.

Considering prostatic hypertrophy in the male patients and menopause in the female patients, the PVRs of the patients aged ≥50 years were compared with those of the patients aged ≤50 years, and for statistical analysis, SPSS ver. 17 for Windows was used. The Mann-Whitney test was used to assess the statistical significance of the following: the difference in PVR between the UTI and non-UTI groups; the difference in PVR between the UTI and non-UTI groups according to sex and age; the number of days from the onset of stroke to admission to the rehabilitation unit; and the number of days of Foley catheter indwelling since the onset of stroke. The difference in stroke type and location between the two groups was assessed via chi-square test (p<0.05).

Of the total of 188 patients, 105 were male and 83 female, with a mean age of 67.1±13.3. The numbers of male patients in the UTI and non-UTI groups were 41 and 64, respectively, and the numbers of female patients in the UTI and non-UTI groups were 33 and 50, respectively. The proportions of the patients with lesions in the anterior cerebral artery in the UTI and non-UTI groups were 8.1 and 7.9%, respectively; in the middle cerebral artery, 56.8 and 59.6%; in the posterior cerebral artery, 5.4 and 5.3%; and in the brainstem, 13.5 and 22.8%. There was thus no significant difference in the locations of the lesions between the two groups. For the stroke types, the proportions of the patients with ischemic stroke in the UTI and non-UTI groups were 43.2 and 89.5%, respectively, and those of the patients with hemorrhagic stroke were 56.8 and 10.5%. There was thus no significant difference between the two groups (Table 1). The PVR and K-MMSE scores, the initial K-MBI scores at the time of admission to the rehabilitation unit, and the Foley catheter indwelling time, however, were significantly different between the two groups (p<0.01) (Table 1). In particular, multivariate regression analysis showed that the initial K-MBI scores at the time of admission to the rehabilitation unit influenced the onset of UTI (Table 2).

For ≥100 and <100 ml PVR, the likelihoods of the onset of UTI were 31% and 68%, respectively (odds ratio: 4.87) (Table 3).

The mean PVR was higher in the UTI group (106.5±102.0 ml) than in the non-UTI group (106.5±102.0 ml) (p<0.01) (Table 1). The mean PVRs in the female patients (n=83) in the UTI and non-UTI groups were 97.0±77.4 and 69.1±114.2 ml (p<0.01), respectively, and in the male patients (n=105), 114.2±118.5 and 41.4±50.4 ml (p<0.01). Thus, the mean PVRs were higher in the UTI group than in the non-UTI group for both the female and male patients (Fig. 1). For the patients aged <50 years (n=31), the mean PVRs in the UTI and non-UTI groups were 138.0±174.6 and 33.9±69.3 ml, respectively, and for the patients aged ≥50 years (n=157), 100.5±81.9 ml (p<0.05) and 62.8±90.2 ml (p<0.01). Thus, the mean PVR was significantly higher in the UTI group than in the non-UTI group irrespective of the patient's age (Fig. 2).